{"harness": {"recipes": 1, "summary": "mini-SWE-agent is a minimal Python SWE agent (package `mini-swe-agent`, src/minisweagent). Its CLI entrypoint is the `mini` console script (minisweagent.run.mini:app, typer), which builds a config from src/minisweagent/config/mini.yaml plus CLI flags, instantiates a LitellmModel (default model class), a LocalEnvironment (runs each bash action as a subprocess in the current working directory) and the InteractiveAgent, then loops: it calls litellm.completion() with a single `bash` function tool, executes the returned tool calls, feeds the JSON-rendered output back as tool messages, and stops when the model runs `echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT`. The model is chosen by MSWEA_MODEL_NAME (or -m); with an `openai/<name>` model name litellm speaks the OpenAI chat-completions API and honors OPENAI_BASE_URL/OPENAI_API_BASE and OPENAI_API_KEY (the config also passes model_kwargs.api_base explicitly). The trajectory (all messages, tool outputs, exit status) is written as JSON to the path given by -o.", "recipe": "swe-agent-mini-swe-agent@04d809ceab9df28f9adaed044884180159172930", "tasks_passed": 7, "harness": "swe-agent-mini-swe-agent", "first_run": "2026-09-23T10:31:10", "domains": ["swe"], "runs": 16, "base_image": "python:3.12-slim-bookworm", "finished": 16, "last_run": "2026-09-25T22:09:23", "repo": "https://github.com/SWE-agent/mini-swe-agent", "passes": 7, "last_run_id": "20260925T220613-swe-agent-mi-b7a243ece02c-polyglot_python_connect", "use_case": "solves GitHub issues and programming tasks in a codebase by executing bash commands in sandboxed environments", "tasksets": ["aider_polyglot", "algotune", "bird-bench", "humanevalfix", "quixbugs", "swebench-verified"], "results": {"bird-bench/card_games__372": {"passes": 1, "last": "2026-09-24T22:55:43", "last_run": "20260924T225451-sw-mini-swe-a-card_games__372", "last_tests": null, "last_outcome": "pass", "last_reward": 1, "runs": 1}, "algotune/algotune-vertex-cover": {"passes": 1, "last": "2026-09-25T22:01:22", "last_run": "20260925T215245-swe-agent-mi-ca91fde3f902-algotune-vertex-cover", "last_tests": {"summary": "2 failed, 1 passed", "total": 0, "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "pass", "last_reward": 1, "runs": 1}, "aider_polyglot/polyglot_go_palindrome-products": {"passes": 0, "last": "2026-09-24T22:43:46", "last_run": "20260924T224300-sw-mini-swe-a-polyglot_go_palindrome-products", "last_tests": null, "last_outcome": "fail", "last_reward": 0, "runs": 1}, "quixbugs/quixbugs-python-pascal": {"passes": 1, "last": "2026-09-24T23:06:16", "last_run": "20260924T230534-swe-agent-mi-quixbugs-python-pascal", "last_tests": {"summary": "6 passed", "total": 6, "passed": 6, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "pass", "last_reward": 1, "runs": 1}, "aider_polyglot/polyglot_python_bowling": {"passes": 1, "last": "2026-09-24T21:58:38", "last_run": "20260925T045731-swe-agent-mi-polyglot_python_bowling", "last_tests": {"summary": "1 error", "total": 0, "cases": null, "aborted": "1 error during collection: fix_test.py", "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "fail", "last_reward": 0, "runs": 2}, "swebench-verified/psf__requests-5414": {"passes": 1, "last": "2026-09-24T22:59:20", "last_run": "20260924T225915-sw-mini-swe-a-psf__requests-5414", "last_tests": {"summary": "131 passed, 1 xfailed, 158 errors", "total": 0, "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "pass", "last_reward": 1, "runs": 2}, "humanevalfix/python-12": {"passes": 1, "last": "2026-09-24T22:54:18", "last_run": "20260924T225327-sw-mini-swe-a-python-12", "last_tests": {"summary": "1 passed", "total": 1, "passed": 1, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "pass", "last_reward": 1, "runs": 1}, "humanevalfix/python-11": {"passes": 1, "last": "2026-09-24T23:08:36", "last_run": "20260924T230803-swe-agent-mi-python-11", "last_tests": {"summary": "1 passed", "total": 1, "passed": 1, "failed": 0, "agent_written": 0, "failed_names": []}, "last_outcome": "pass", "last_reward": 1, "runs": 1}, "aider_polyglot/polyglot_python_connect": {"passes": 0, "last": "2026-09-25T22:09:23", "last_run": "20260925T220613-swe-agent-mi-b7a243ece02c-polyglot_python_connect", "last_tests": {"summary": "5 failed, 5 passed", "total": 10, "passed": 5, "failed": 5, "agent_written": 0, "failed_names": ["test_an_empty_board_has_no_winner", "test_illegal_diagonal_does_not_make_a_winner", "test_nobody_wins_crossing_adjacent_angles", "test_o_wins_crossing_from_top_to_bottom", "test_only_edges_does_not_make_a_winner"]}, "last_outcome": "fail", "last_reward": 0, "runs": 2}}, "models": ["bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "bedrock/us.anthropic.claude-sonnet-4-5-20250929-v1:0"], "tasks_tried": 9, "api_style": "openai", "commit": "04d809ceab9df28f9adaed044884180159172930"}, "profile": {"evidence": "README.md describes it as solving GitHub issues using only bash (lines 5-55); agent/default.py shows the core loop executes bash commands via environments (lines 126-157); config files (default.yaml, mini.yaml) contain prompts instructing the LM to edit files, run tests, and analyze code using bash; local.py and docker.py show environments execute bash via subprocess; designed for SWE-bench benchmark evaluation (README line 28)", "harness": "swe-agent-mini-swe-agent", "domains": ["swe"], "source": "claude -p", "cost_usd": 0.26937315, "capabilities": ["edits-files", "runs-shell", "runs-tests", "reads-docs", "long-horizon"], "seconds": 73, "languages": ["python"], "at": "2026-09-25T22:04:18", "use_case": "solves GitHub issues and programming tasks in a codebase by executing bash commands in sandboxed environments", "not_for": ["no web browsing", "no SQL queries", "no API calls to external services", "no GUI automation"], "commit": "04d809ceab9df28f9adaed044884180159172930"}, "recommendations": {"model": "haiku", "at": "2026-09-25T22:13:49", "profile_source": "claude -p", "harness": "swe-agent-mini-swe-agent", "recs": [{"language": null, "task": "sphinx-doc__sphinx-10323", "why": "It's a real GitHub issue reflecting the harness's primary use case\u2014validates whether it can solve production problems beyond toy tasks.", "domain": "swe", "taskset": "swebench-verified", "score": null}, {"score": null, "domain": "swe", "taskset": "aider_polyglot", "why": "Same algorithm as the Python bowling task (which has mixed results), but in a memory-safe compiled language\u2014shows if the harness generalizes across languages or hits Rust-specific friction.", "language": "rust", "task": "polyglot_rust_bowling"}, {"task": "polyglot_cpp_allergies", "score": null, "why": "New language and new algorithm; tests whether the harness can handle C++ compilation, linking, and memory semantics.", "taskset": "aider_polyglot", "domain": "swe", "language": "cpp"}, {"why": "Java version of a bug-fix pattern it solved in Python (pascal)\u2014measures if one-line debugging skills transfer or if each language requires relearning.", "domain": "swe", "taskset": "quixbugs", "score": null, "task": "quixbugs-java-flatten", "language": "java"}], "source": "llm", "based_on_run": "20260925T220613-swe-agent-mi-b7a243ece02c-polyglot_python_connect", "cost_usd": 0.057205000000000006}, "runs": [{"run": "20260925T220613-swe-agent-mi-b7a243ece02c-polyglot_python_connect", "started": "2026-09-25T22:09:23", "finished": "2026-09-25T22:12:34", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "aider_polyglot", "name": "polyglot_python_connect"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 0, "verifier_rc": 0, "tests": {"summary": "5 failed, 5 passed", "total": 10, "passed": 5, "failed": 5, "agent_written": 0, "failed_names": ["test_an_empty_board_has_no_winner", "test_illegal_diagonal_does_not_make_a_winner", "test_nobody_wins_crossing_adjacent_angles", "test_o_wins_crossing_from_top_to_bottom", "test_only_edges_does_not_make_a_winner"]}, "calls": 39, "seconds": 185, "input_tokens": 709619, "output_tokens": 24136, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 0 \u00b7 5 failed, 5 passed"}, {"run": "20260925T215245-swe-agent-mi-ca91fde3f902-algotune-vertex-cover", "started": "2026-09-25T22:01:22", "finished": "2026-09-25T22:02:59", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "algotune", "name": "algotune-vertex-cover"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 1, "tests": {"summary": "2 failed, 1 passed", "total": 0, "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 19, "seconds": 93, "input_tokens": 261205, "output_tokens": 12287, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 2 failed, 1 passed \u00b7 verifier exited 1"}, {"run": "20260924T232628-swe-agent-mi-polyglot_python_connect", "started": "2026-09-24T23:27:06", "finished": "2026-09-24T23:29:38", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "aider_polyglot", "name": "polyglot_python_connect"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 0, "verifier_rc": 0, "tests": {"summary": "6 failed, 4 passed", "total": 10, "passed": 4, "failed": 6, "agent_written": 0, "failed_names": ["test_an_empty_board_has_no_winner", "test_illegal_diagonal_does_not_make_a_winner", "test_nobody_wins_crossing_adjacent_angles", "test_o_wins_crossing_from_top_to_bottom", "test_only_edges_does_not_make_a_winner", "test_x_wins_using_a_convoluted_path"]}, "calls": 38, "seconds": 147, "input_tokens": 588071, "output_tokens": 17914, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 0 \u00b7 6 failed, 4 passed"}, {"run": "20260924T230803-swe-agent-mi-python-11", "started": "2026-09-24T23:08:36", "finished": "2026-09-24T23:09:08", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "humanevalfix", "name": "python-11"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 0, "tests": {"summary": "1 passed", "total": 1, "passed": 1, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 12, "seconds": 29, "input_tokens": 46368, "output_tokens": 2749, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 1 passed"}, {"run": "20260924T230534-swe-agent-mi-quixbugs-python-pascal", "started": "2026-09-24T23:06:16", "finished": "2026-09-24T23:06:54", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "quixbugs", "name": "quixbugs-python-pascal"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 0, "tests": {"summary": "6 passed", "total": 6, "passed": 6, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 8, "seconds": 32, "input_tokens": 33930, "output_tokens": 3750, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 6 passed"}, {"run": "20260924T225915-sw-mini-swe-a-psf__requests-5414", "started": "2026-09-24T22:59:20", "finished": "2026-09-24T23:04:23", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "swebench-verified", "name": "psf__requests-5414"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 0, "tests": {"summary": "131 passed, 1 xfailed, 158 errors", "total": 0, "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 100, "seconds": 264, "input_tokens": 2311813, "output_tokens": 20182, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 131 passed, 1 xfailed, 158 errors"}, {"run": "20260924T225451-sw-mini-swe-a-card_games__372", "started": "2026-09-24T22:55:43", "finished": "2026-09-24T22:56:05", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "bird-bench", "name": "card_games__372"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 0, "tests": null, "calls": 7, "seconds": 17, "input_tokens": 21923, "output_tokens": 722, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1"}, {"run": "20260924T225327-sw-mini-swe-a-python-12", "started": "2026-09-24T22:54:18", "finished": "2026-09-24T22:54:50", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "humanevalfix", "name": "python-12"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 1, "verifier_rc": 0, "tests": {"summary": "1 passed", "total": 1, "passed": 1, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 8, "seconds": 21, "input_tokens": 20477, "output_tokens": 1295, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 1 passed"}, {"run": "20260924T224628-sw-mini-swe-a-psf__requests-5414", "started": "2026-09-24T22:47:46", "finished": "2026-09-24T22:53:25", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "swebench-verified", "name": "psf__requests-5414"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 0, "verifier_rc": 1, "tests": {"summary": "15 failed, 116 passed, 1 xfailed, 158 errors", "total": 0, "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 55, "seconds": 167, "input_tokens": 857570, "output_tokens": 12252, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 0 \u00b7 15 failed, 116 passed, 1 xfailed, 158 errors \u00b7 verifier exited 1"}, {"run": "20260924T224300-sw-mini-swe-a-polyglot_go_palindrome-products", "started": "2026-09-24T22:43:46", "finished": "2026-09-24T22:46:26", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "aider_polyglot", "name": "polyglot_go_palindrome-products"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 0, "verifier_rc": 0, "tests": null, "calls": 37, "seconds": 150, "input_tokens": 354082, "output_tokens": 10572, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 0"}, {"run": "20260925T045731-swe-agent-mi-polyglot_python_bowling", "started": "2026-09-24T21:58:38", "finished": "2026-09-24T22:01:23", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "aider_polyglot", "name": "polyglot_python_bowling"}, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": 0, "verifier_rc": 0, "tests": {"summary": "1 error", "total": 0, "cases": null, "aborted": "1 error during collection: fix_test.py", "passed": 0, "failed": 0, "agent_written": 0, "failed_names": []}, "calls": 32, "seconds": 160, "input_tokens": 503026, "output_tokens": 20716, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 0 \u00b7 tests stopped: 1 error during collection: fix_test.py"}, {"run": "ic-record", "started": "2026-09-23T18:53:53", "finished": "2026-09-23T18:54:06", "status": "done", "kind": "prompt", "harness": "swe-agent-mini-swe-agent", "task": null, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": null, "verifier_rc": null, "tests": null, "calls": 4, "seconds": 12, "input_tokens": 6935, "output_tokens": 702, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "error", "verifier_says": "no reward written"}, {"run": "routes-smoke", "started": "2026-09-23T18:19:24", "finished": "2026-09-23T18:19:36", "status": "done", "kind": "prompt", "harness": "swe-agent-mini-swe-agent", "task": null, "model": "bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0", "reward": null, "verifier_rc": null, "tests": null, "calls": 4, "seconds": 9, "input_tokens": 6299, "output_tokens": 438, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "error", "verifier_says": "no reward written"}, {"run": "20260923T104631", "started": "2026-09-23T10:46:37", "finished": "2026-09-23T10:47:00", "status": "done", "kind": "prompt", "harness": "swe-agent-mini-swe-agent", "task": null, "model": "bedrock/us.anthropic.claude-sonnet-4-5-20250929-v1:0", "reward": null, "verifier_rc": null, "tests": null, "calls": 6, "seconds": 22, "input_tokens": 12101, "output_tokens": 916, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "error", "verifier_says": "no reward written"}, {"run": "20260923T103337-polyglot_python_bowling", "started": "2026-09-23T10:33:39", "finished": "2026-09-23T10:36:50", "status": "done", "kind": "harbor", "harness": "swe-agent-mini-swe-agent", "task": {"taskset": "aider_polyglot", "name": "polyglot_python_bowling"}, "model": "bedrock/us.anthropic.claude-sonnet-4-5-20250929-v1:0", "reward": 1, "verifier_rc": 0, "tests": {"summary": "69 passed", "total": 31, "passed": 31, "failed": 0, "agent_written": 38, "failed_names": []}, "calls": 30, "seconds": 187, "input_tokens": 318418, "output_tokens": 13013, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "scored", "verifier_says": "reward 1 \u00b7 69 passed"}, {"run": "20260923T103035", "started": "2026-09-23T10:31:10", "finished": "2026-09-23T10:31:40", "status": "done", "kind": "prompt", "harness": "swe-agent-mini-swe-agent", "task": null, "model": "bedrock/us.anthropic.claude-sonnet-4-5-20250929-v1:0", "reward": null, "verifier_rc": null, "tests": null, "calls": 7, "seconds": 29, "input_tokens": 16488, "output_tokens": 1511, "errors": 0, "last_action": "bash: echo COMPLETE_TASK_AND_SUBMIT_FINAL_OUTPUT", "outcome": "error", "verifier_says": "no reward written"}]}